No excuses for missing work: AI identifies people with the flu by their tone of voice on the phone

Without a doubt, one of the main excuses for missing appointments is the flu. Even if this isn’t your case, a close friend or family member has certainly used it, or you know someone who has told this lie to miss an important appointment.

However, advances in technology have directed studies in broad areas, including voice recognition, and this will have implications for several situations. Learn how voice recognition works Artificial Intelligence that identifies people with the flu.

New flu detector

Now it won’t be so easy to lie about having the flu! That’s because a group of scientists have developed a Artificial intelligence that recognizes the voices of people with the flu.

Obviously, the main goal of this technology is not to identify lying employees, but to reduce the number of people seeking medical care. Initially, the equipment is intended to detect signs and symptoms of flu, but in the future, the idea is to expand its use to other health conditions.

Experts explain that the investigation is carried out through the harmonic tones of our voice, which decrease in amplitude as the frequency increases. When we are infected by a virus, this pattern of voice tones changes.

The results of the project were found from the analysis of approximately 630 people, of which 111 were diagnosed with flu confirmed by a doctor.

During the call, a few commands were given. The first was for participants to count from 1 to 40. Then, they were asked to tell a story. At this point, all the algorithms included in the system contained the information necessary for the machine to understand the difference between the tones of voice with and without the flu.

The collection process is still ongoing, but with some results being analyzed, researchers are already very happy with the system. This is because it has already shown significant results, since Artificial Intelligence has so far detected 70% of the cases, which is great for the testing phase.